Page History

y are an independent 

Remarks on the Extended Range

...

The Extended Range (monthly) products cover the period up to Day46.  They are derived from 100 member ensemble with resolution 36km, independent of ENS, run daily.   They are derived from an extension of the normal 15-day ENS twice per week on Mondays and Thursdays.  This is a time scale lying between:

• medium range forecasts (ENS to day15).  These are mainly governed by atmospheric initial values (background plus new observed data) but less so on ocean temperature information.
• seasonal forecasts. These are more reliant on predictability of the oceans and on the impact that tropical sea-surface temperatures have on the atmospheric circulation.  However some initial  However some initial atmospheric data is carried over from the medium ranges.

This is a particularly difficult time range as it is:

...

For all forecasts the ocean, atmosphere and land surface are initialised to be as close to reality as possible.  Then the coupled models calculate the evolution of the atmospheric and oceanic systems.  No "artificial" terms are introduced to try to reduce the drift of the model.  No steps are taken to remove or reduce any imbalances in the coupled model initial state.  The effect of model drift can be estimated from re-forecasts and thus may be "removed" from the latest model solution during the post-processing.  This is not a perfect solution, but rather an expedient one.  Model drift characteristics may also depend somewhat on the prevailing synoptic pattern, and will not be accounted for fully.

The extended range model climate (ER-M-climate) drifts towards becoming rather too cold at longer lead-times in wintertime high latitudes.  Hence the .  ???.    Hence the anomaly in forecast temperatures against ER-M-climate temperatures may be too large.  The  The magnitude of the drift is not uniform.  At longer lead-times the trend in northern China is towards colder values but less so in Siberia and Canada.  The variation may be due to the analysed initial snowpack conditions and/or snowmelt in marginal snow cover areas in these areas.  Issues regarding this are being addressed with a multi-layer snow scheme currently being developed.  

After about 10 days of forecasts, the spread of the ensemble can become very large.  A significant shift can be detected by comparing probability distribution functions of the latest model and the ER-M-climate.

...